Building IoT Systems Modeling: A Object-oriented Metamodeling Approach.

Buildings are supersystems consisting of many constituent systems that require coordination and interaction to function coherently. Modeling buildings and representing them in both readable forms by humans and computers is important for efficient coordination, simulation, control, and optimization of the systems within buildings and for the integration and creation of other novel applications and functions. Most commercial buildings come with a Building Energy Management System (BEMS), which has representations of aspects of a building and its systems. These representations in the BEMSs, however, do not wholly model a building (and its spatial entities), its systems and subsystems, and their relationships. This is because their primary function is to control heating, ventilation, air-conditioning (HVAC), and lighting. Although there are many studies on modeling the built environment for operational purposes, recently, Project Haystack and Brick have made good progress by adopting a metadata approach to modeling and providing readable representations of buildings and their systems. However, the tags and tagset systems have some limitations that stem from the lack of structure in defining entities, their properties, and relationships. In this study, we identify five problems with the metadata approach to modeling the built environment and attempt to address these problems with an object-oriented metamodel: Energy Things Metamodel (ETM), that fully models a building, its systems, subsystems, and their relationships. ETM maintains a consistent naming convention of entities with other projects like (Project Haystack and Brick) and uses an object-oriented approach to modeling, producing a metamodel that best represents a building and its systems. In this study, a UML implementation of the core structure of ETM is achieved.